It has become increasingly commonplace to use grids of multiple node devices to perform widely varied analyses of large data sets (e.g., what is commonly referred to as “big data”) of widely varied types. Such grids of nodes often receive such data sets from one or more grids of storage devices at which such data sets may be assembled in what may be a queue of data sets awaiting analysis. Such a queue of data sets may be formed among the one or more grids of storage devices in an effort to cause a grid of node devices to be as fully utilized as possible.